In vitro study
Study design
One pint (473.716 mL) of pineapple juice placed in a glass container was concentrated by a microwave heating evaporation process performed in a domestic microwave oven (MS11K3000AS, Samsung, Maetan-dong, Yeongtong District, Suwon, South Korea). The microwave’s power was set to 850 W for 25 min. Five grams of modified potato starch (Gel’M instantané ed. végétal, Nutrisens MEDICAL, Francheville, France) for every 40 mL of concentrated pineapple juice were added to the concentrated pineapple juice in order to thicken the solution (Fig. 1).
To evaluate the reproducibility of the microwave heating evaporation processes, we compared five 10-mL syringes filled with concentrated pineapple juice plus modified potato starch obtained from five consecutive and independent concentration process (Fig. 2).
To determine whether the solution so obtained retained the paramagnetic characteristics required by magnetic resonance imaging (MRI) contrast agent, we compared it to a contrast agent we routinely use for MRA (Fig. 3). We scanned three 10-mL syringes filled with: (1) pineapple juice; (2) concentrated pineapple juice plus modified potato starch; and (3) saline diluted gadoteridol (279.3 mg/mL, ProHance®, Bracco Altana Pharma, Constance, Germany) at a 3% concentration. This concentration was chosen based on the hypothesis that the contrast agent, after being injected at a flow rate of 2.5 mL/s, is homogeneously mixed with the blood pool on its arrival to the heart; therefore, for a mean cardiac output of 5 L/min (83.3 mL/s), its concentration is about 2.5/83.3 = 3%.
MRA protocol and image analysis
MRA was performed with a 1.5-T scanner (Achieva, version 2.6, Philips Medical Systems, Eindhoven, The Netherlands) using a 32-channel body phased-array coil. A three-dimensional (3D) spoiled gradient-echo sequence was acquired in the axial plane, with the following technical parameters: repetition time 3.3 ms; echo time 1.2 ms; flip angle 20°; right to left phase encoding; 512 × 512 matrix with an isotropic voxel of 1.5 mm3.
All MRA studies were analysed in consensus by two experienced observers with > 5 years of experience in cardiovascular MRI using OsiriX MD DICOM Viewer (Pixmeo Inc., Bernex, Switzerland).
To compare the intensity of the scanned syringe, a region of interest with a mean size of 1 cm2 was placed in the axial plane in the centre of each image (Figs. 2 and 3).
In vivo study
Study design
The study was piloted in agreement with the 1964 Helsinki declaration and its later amendments and was approved by the ethics committee of our institution. Before MRA, all patients herein considered were informed about the possible use of their data for study purposes and gave their consent. Patient information was anonymised before the analysis.
Inclusion criteria were: symptomatic AF refractory to at least one anti-arrhythmic drug; age ≥ 18 years; and preserved left ventricular ejection fraction at echocardiography. Exclusion criteria were: active hyperthyroidism; impaired left ventricular function; pregnancy; previous oesophageal-gastric surgery; and contraindications to anticoagulation or MRA.
MRA protocol and image analysis
For the in vivo study, the same not electrocardiographically gated, free-breath sequence was performed after intravenous injection of 0.1 mmol/kg of gadoteridol at a rate of 2.5 mL/s, followed by a 20-mL saline bolus at the same rate. The mean sequence time was 25 s (range 19–32 s). Bolus tracking was used to start the sequence at the exact moment the contrast intensified during the venous phase of pulmonary circulation, to guarantee the maximum signal intensity in the pulmonary veins and into the LA. The oesophagus was intensified by administration of 40 mL of concentrate pineapple with 5 g of modified potato starch (Gel’M instantané ed. végétal, Nutrisens MEDICAL, Francheville, France), served with a disposable plastic spoon, while the patients were on the scanning table before the sequence acquisition.
After MRA, all patients were clinically monitored for 30 min. Any adverse effects or anomalies were registered after RFCA before discharge.
For the in vivo study, the feasibility of LA volume and left appendage volume calculation and the quality (classified as diagnostic or non-diagnostic) of 3D maximum intensity projection and 3D volume rendering reconstructions were evaluated to assess the spatial position of the oesophagus and identify appendage morphology and any anatomical variations of pulmonary veins.